OPA1656: High-Performance CMOS Audio Op Amp

[MarcelvdG]

It could be useful to also read this thread about stability issues and solutions: https://www.diyaudio.com/community/...plit-from-opa1656-thread.377331/#post-6785883

 

[Wydra9]

Thank you very much for the post, I will read it for sure. OPA828 has better parameters than OPA1641 and OPA1655 but maybe these opamps sound better?And here is my PCB from RIAA preamplifier

 

[Ken Newton]

I just noticed that there appears to be a typo error in the OPA1656 data sheet, in regards to the specification for Unity gain bandwidth. It’s oddly shown as 20MHz, instead of the expected 53MHz. Not a big deal, but I wonder where the 20MHz figure came from.

 

[Mark Johnson]

Like many opamps, the OPA1656 does NOT have constant gain-bandwidth product. There are additional poles and zeroes in its frequency response, which introduce additional lumpity bumpity behaviors in the gain-vs-frequency and phase-vs-frequency plots. Take a look at this figure from its datasheet.

Gain at 1 kHz is approx 95 dB (56000 x) so the GBW product at 1 kHz is approx 56 million, namely, 56 MHz.


_

 

[DSP_Geek]

That open-loop gain & phase graph was done with 10 pf load, maybe 10 cm of audio coax -- ignoring the connector. The spec sheet overshoot graphs show you want care with capacitive loads such as lines longer than an inch or two, and reading between the lines suggests best practice is a 50-ish ohm isolation resistor immediately after the opamp feedback loop.

 

[jan.didden]

Yes, and the difference between 20MHz and 50MHz seems like a whole lot but it's just a bit more than an octave.

Jan

 

[pixelgarden]

Hello everyone. I upgraded my OPamp on my Aiyima A07 from stock NE5532 with OPA 1656, having read alot of good things about them. But i am very dissappointed. The sparkling high end details dissappeared while making the whole presentation very dull & boring. The audio feels much better overall with NE 5532. How is this possible. Am i doing something wrong ? Or are the 1656 chip supposed to have a burn-in time ?

 

[MarcelvdG]

Depending on the circuit and layout, it might be oscillating and distorting a lot. Do you have a schematic you can post?

 

[Markw4]

Sometimes 'sparkling high end details' are actually distortion. Since this is a class-D amp there may be some switching noise getting into the analog circuitry. Do you have a scope?

 

[pixelgarden]

Depending on the circuit and layout, it might be oscillating and distorting a lot. Do you have a schematic you can post?

I don't have the schematics but lots of other users of this amp has reported great enhancement in SQ after the upgarde. Also, I have used 4.7uF electrolytic + 0.22 uF film cap across the OPamp power pins. But removing them doesn't improve the sound either !

 

[pixelgarden]

Sometimes 'sparkling high end details' are actually distortion. Since this is a class-D amp there may be some switching noise getting into the analog circuitry. Do you have a scope?

Is it ? It feels very clean & lively with the NE 5532 though. I don't have the scope but the measurments are here on this page : https://www.audiosciencereview.com/...ds/aiyima-a07-tpa3255-review-amplifier.18984/
Can I use one NE 5532 & OPA 1656 simultaneously in each seperate channel just for trying ?

 

[Markw4]

Would be curious to see if trying MUSES8920 might help some. Basically though wouldn't be too surprising if this unit has switching noise problems that might in a round about way explain what you are hearing.

 

[pixelgarden]

Unfortunately I have just the basic NE 5532 & OPA 1656 that i purchased recently. It's just an hour old. Do these need burn in time to open up ? It's very contrained in presentation with OPA 1656 & most of the music seems to come from middle of the two channel speaker system, while with 5532 it seems to play with a larger soundstage like all other amps.

 

[DSP_Geek]

Look to my post above for a possibility: the opamp could be driving a capacitive load hefty enough to destabilise it.

 

[pixelgarden]

Look to my post above for a possibility: the opamp could be driving a capacitive load hefty enough to destabilise it.

Okay.

 

[stocktrader200]

That 1656 looks outstanding, like an LF353 on steroids. Unity gain stable, JFet inputs , 150 dB open loop gain. It should be close to perfect on a properly laid out PCB.

 

[abraxalito]

JFet inputs ,

Actually MOS, not junction FETs. This isn't merely an academic difference, check the noise behaviour of 1656 (top) vs 1642 (bottom).

 

[Ghoostknight]

i really liked the opa1656 while op amp rolling, are there newer opamps out yet that rival the 1656?

 

[bramsat]

I got the aiyima A07 with the opa 1676 from aliexpress...the mids and hights are better.
https://nl.aliexpress.com/item/1005...st_main.4.503a79d2gHTUJ3&gatewayAdapt=glo2nld

 

[turion64]

There is no OPA1676, is there? Maybe 1678. Or did you mean 1656 as is referred to in the link?